Workshop: Reference Image Datasets for Medical Image ... Workshop: Reference Image Datasets for Medical…

  • Published on
    27-Jul-2018

  • View
    213

  • Download
    0

Transcript

<ul><li><p>1</p><p>Workshop: Reference Image Datasets for Medical Image Processing Research and Development</p><p>August 27, 2002, 16:10-17:30</p><p>Alexander HorschChair of WG MIP</p><p>Topic summary MIP is of increasing importance for modern medicine</p><p>(CAD, CAS, etc) However: Results of R&amp;D work are usually not or hardly </p><p>comparable due to the fact that each R&amp;D group uses its own datasets for to develop and evaluate their methods</p><p> Furthermore: datasets themselves are often insufficient or too small for clinically sound evaluation studies</p><p> There is an urgent need to establish, step by step, a reference image database providing numerous high-quality datasets for medical image processing R&amp;D groups from different domains. </p><p> In September 2001, with the EFMI WG an initiative has been started to begin this work.</p></li><li><p>2</p><p>Goal of the workshop During the workshop, the existing concept and roadmap for to </p><p>create such a database, as well as first available datasets shall be presented and discussed. </p><p> In the end of the workshop, a list of recommendations for a revision of both concept and roadmap shall be agreed on.</p><p> Furthermore, interested colleagues can join the efforts of EFMI WG MIP.</p><p>Professional interests of audience</p><p> intended audience are colleagues working in the field of medical image processing, either on methodological research or on development of clinical applications like CAD or CAS. Possible motivations for to join the workshop can be: to get information about existing material and how to get it </p><p>for the own work identification of possible own contributions to the database for to get up-to-date information about the subject for to meet experts in this field</p></li><li><p>3</p><p>Expected outcome A list of recommendations on further development </p><p>of reference database New contacts between the WG and potential </p><p>participants in the development</p><p>Agenda Presentation of </p><p> Background Concept Roadmap First Datasets</p><p> Presentation of Casim@ge (Christian Lovis, Geneva) Discussion of the presented material Collection (brainstorming) and sorting of ideas for </p><p>improvement Selection and final agreement on a list of </p><p>recommendations</p></li><li><p>4</p><p>LearningLearning</p><p>Image -recordingImage -</p><p>recordingPre-</p><p>processingPre-</p><p>processing SegmentationSegmentationParameter-extraction</p><p>Parameter-extraction ClassificationClassification</p><p>ResultsResultsSceneScene ImageImage ImageImage ObjectsObjects FeaturesFeatures</p><p> MIP Teaching</p><p> Calibration Modality Simulators Standardized</p><p>- acquisition- documentation of</p><p>acquisition chains- paradigm / instruction- image format</p><p> Image improvement Image Registration Image econstruction Image/Data Fusion 3D Volume</p><p>(Re)construction Visualization Detection of ROIs Mapping VR Augmented Reality Image reconstruction from partial cquisition</p><p> Segmentation</p><p> Object detection &amp; counting Quantification/ Measurement Identification of anatomic structures Visualization Reference Validation</p><p> CAD Pattern recognition Detection of location Fibre tracking from diffusion MRI</p><p> Storage &amp; documentation Image communication (telemedicine)</p><p> CBIR Compression (lossy / non-lossy) Cryptography / Watermarking</p><p>The MIP Specialty</p><p>Background MIP is growing interdisciplinary field Basic research, clinical evaluation, algorithm design, clinical studies Own image datasets by almost each R&amp;D group Algorithms tend to be optimized to the one or few acquisition units </p><p>available Comparison of research results from different R&amp;D groups rarely </p><p>possible Datasets usually do not represent the medical variety needed for large </p><p>clinical studies Solution multi-center studies usually suffers from unbalanced usage of </p><p>different imaging units</p></li><li><p>5</p><p>EFMI WG MIP Initiated 09/2001 London (Medinfo) Mission:</p><p> Foster discussion and activities on theory and practice of MIP comprising: integration of decision support by MIP into clinical practice</p><p>(including clinical evaluation, standardization, technology transfer) Work products:</p><p> RID-MIP MIP R&amp;D Groups IS WG Website</p><p>WG Members Austria: Prinz, Vogl Finland: Sipil Germany: Horsch, Wittenberg, Schneider, Zahlmann Switzerland: Valle The Netherlands: Verdonck</p><p>Interested colleagues from further EFMI countries: Spain: Gomez Vilda</p><p>coming from universities, Siemens, hospitals.</p></li><li><p>6</p><p>Agenda Presentation of </p><p> Background Concept Roadmap First Datasets</p><p> Presentation of Casim@ge (Christian Lovis, Geneva) Discussion of the presented material Collection (brainstorming) and sorting of ideas for </p><p>improvement Selection and final agreement on a list of </p><p>recommendations</p><p>Overall Approach RIDSs (well-defined, validated) from experienced providers RIDSs meet quality criteria defined by interdisciplinary expert groups </p><p>(domain-specific) Bilateral agreements / contracts between provider &amp; users of a RIDS EFMI WG MIP </p><p> proposes tasks / challenges supports RIDSs specifications supports RIDSs dissemination organizes special sessions at conferences evaluates overall impact</p></li><li><p>7</p><p>Criteria for Task Assessment Relevance of medical problem in terms of</p><p> Mortality &amp; morbidity statistics DRG statistics (e.g. USA, Australia, Austria) What is funded by WHO, EC, Current science activities, Literature</p><p> Role of imaging in the diagnostic / therapeutic process key role, important role, marginal role</p><p> Potential for improvement of outcome by improvement of image reading / image usage / image analysis life years, cost savings</p><p> MIP task paths to find solutions exists MIP challenge no real ideas present of how to find a solution</p><p>Leading causes of death, USA 1999</p><p>From EFMI-WG-MIP-2310-00.10 draft 2002-07-24 Fig. 1: Part of Table 1: Deaths, percent of total deaths, and death rates for the 10 leading causes of death in selected age groups, by race and sex: United States, 1999. National Vital Statistics Reports, Vol. 49, Nr. 11, October 12, 2001, page 14</p></li><li><p>8</p><p>Medline Meta-Studies</p><p>TumorstadieneinteilungLymphatische Metastasierung4</p><p>PrognoseRezidiv4</p><p>EchokardiographieHerzkrankheiten5</p><p>Cholangiopankreatographie,retrograde_endoskopischePankreatitis5</p><p>ComputertomographiePulmonale_Embolie5</p><p>HysterosalpingographieInfertilitt,weibliche5</p><p>BehandlungsergebnisMyokardinfarkt6</p><p>ComputertomographieKoronarkrankheit6</p><p>Einzelphotonen-EmissionscomputertomographieKoronarkrankheit6</p><p>ElektrokardiographieKoronarkrankheit6</p><p>KoronarangiographieRezidiv6</p><p>PrognoseKoronarkrankheit8</p><p>BelastungstestKoronarkrankheit9</p><p>EchokardiographieKoronarkrankheit9</p><p>KoronarangiographieMyokardinfarkt11</p><p>ReihenuntersuchungMammatumoren,Mensch16</p><p>KoronarangiographieKoronarkrankheit17</p><p>MammographieMammatumoren,Mensch25</p><p>ProzedureDiagnosis#</p><p>From EFMI-WG-MIP-2310-00.10 draft 2002-07-24 Table 1: Total number of meta-studies on diagnostic imaging referenced in Medline database on June 5, 2002, sorted by frequency</p><p>Task/challenge proposals CAD for support of breast cancer screening:</p><p>Detect micro-calcifications on all images of the mammography dataset D1 with a 10-fold cross-validated sensitivity &gt; 0.9 at a rate of false positive ROI detections &lt; 0.5! CHALLENGE</p><p> CAD for esophagus tumor staging:Make a robust and fully automatic T1 to T3 staging of esophageal tumors in the EUS still images dataset D2 with an overall classification quality of &gt; 0.6! CHALLENGE</p></li><li><p>9</p><p>Objectives of RID-MIP Provide a methodology for building a medical image database Provide information on existing datasets and who is the provider Provide templates for bilateral agreements / contracts between </p><p>provider and users Provide different kinds of datasets for evaluation of MIP </p><p>methods and applications Follow-up the usage of datasets in terms of published results Bring the prototypical RID-MIP into long-term operation (e.g. at a </p><p>publishing company)</p><p>Addressees of RID-MIP European MIP R&amp;D groups</p><p> MIP R&amp;D groups outside Europe</p><p> Public sector (usually free of charge)</p><p> Private / commercial sector (usually by licensing)</p></li><li><p>10</p><p>MIP system development and the different kinds of datasets</p><p>Modality Simulator Modality Simulator </p><p>PhantomPhantom</p><p>ElktestElktest</p><p>System TrialsSystem Trials</p><p>Development time axis</p><p>Clinical Multicenter Study</p><p>Clinical Multicenter Study</p><p>PatientInformationExamination Information</p><p>--- (x)--- x x</p><p>x (x)Parameters x x</p><p>(x) x x xReference generated</p><p>PtD MSD ETD STD</p><p>Licensing Policy ETDs method-oriented free (agreement) STDs research public sector free, agreement STDs product-oriented license fees user to provider, </p><p>contract EFMI WG MIP will provide templates for agreements &amp; </p><p>contracts</p></li><li><p>11</p><p>Provider/User Agreement 1 Subject</p><p>... timely limited, non-exclusive usage license for the dataset</p><p> 2 User license user is allowed to use the dataset in the framework of the </p><p>following scientific studies: </p><p> 3 Binding to purpose 4 Validity timespan 5 Reference character of the dataset 6 Usage by a third party</p><p>Provider/User Agreement 7 Media and copies 8 Reference copy 9 Fees 10 Co-operation 11 Publication of the results 12 Runtime and irregular termination</p></li><li><p>12</p><p>Organisation &amp; Roles EFMI WG MIP (RID-MIP Project Group)</p><p> initiating, supporting, organizing, assessing</p><p> Technical Board of Experts specifying</p><p> Medical Board of Experts recommending</p><p> Dataset Provider preparing, submitting, offering</p><p> Dataset User using, reporting / publishing</p><p>Agenda Presentation of </p><p> Background Concept Roadmap First Datasets</p><p> Presentation of Casim@ge (Christian Lovis, Geneva) Discussion of the presented material Collection (brainstorming) and sorting of ideas for </p><p>improvement Selection and final agreement on a list of </p><p>recommendations</p></li><li><p>13</p><p>Roadmap (preliminary) Identification and specification of important tasks Selection of 3 tasks from different domains</p><p>where data are probably easy to get Specification and preparation of related ETDs Publishing of task descriptions on the Web </p><p>&amp; dissemination of ETDs Evaluation of impact on R&amp;D</p><p>Time Schedule for phase 1(preliminary, updated August 2002, not yet passed in the WG)</p><p> 1st WG meeting June 6-7, Munich choice &amp; specification of first ETDs until Aug 2002 [early application for funding] preparation of ETDs Oct 2002 to Feb 2003 3rd WG meeting Feb/Mar 2003 dissemination Apr to Aug 2003 [late application for funding] evaluation (3 years)</p></li><li><p>14</p><p>Agenda Presentation of </p><p> Background Concept Roadmap First Datasets</p><p> Presentation of Casim@ge (Christian Lovis, Geneva) Discussion of the presented material Collection (brainstorming) and sorting of ideas for </p><p>improvement Selection and final agreement on a list of </p><p>recommendations</p><p>First datasets Munich group: </p><p> Skin Surface Microscopy of Melanocytic Lesions</p><p> EUS Gastrointestinal tumors</p><p> Erlangen group: MRI of Stroke Lesions</p></li><li><p>15</p><p>D-SSM-ML-300-V1.0 (Dec 2001) Dermatology Skin Surface Microscopy Melanocytic Lesions 150 melanomas 150 naevi In total 300 images</p><p>Standardized RIDS naming</p><p> medical domain Dermatology = D modality Skin Surface Microscopy = SSM subject of imaging Melanocytic Lesions = ML number of images/series 300 version number V1.0 optional: date of issue (Dec2001)</p><p>Dermatology Skin Surface Microscopy Melanocytic Lesions 300D-SSM-ML-300-V1.0</p></li><li><p>16</p><p>Agenda Presentation of </p><p> Background Concept Roadmap First Datasets</p><p> Presentation of Casim@ge (Christian Lovis Geneva) Discussion of the presented material Collection (brainstorming) and sorting of ideas for </p><p>improvement Selection and final agreement on a list of </p><p>recommendations</p><p>Agenda Presentation of </p><p> Background Concept Roadmap First Datasets</p><p> Presentation of Casim@ge (Christian Lovis Geneva) Discussion of the presented material Collection (brainstorming) and sorting of ideas for </p><p>improvement Selection and final agreement on a list of </p><p>recommendations</p></li></ul>

Recommended

View more >